News
Meta Launches New AI System in Poland to Detect Fake Ads
Meta is rolling out an upgraded AI system in Poland to detect fraudulent ads and introducing mandatory identity verification for financial advertisers. The move follows Deputy Prime Minister Krzysztof Gawkowski's request for the European Commission to impose a billion-zloty fine on the company.
Contents
Meta has announced the launch of an advanced artificial intelligence system in Poland to proactively detect fraudulent ads, and is introducing mandatory identity verification for all financial-sector advertisers targeting Polish audiences. The decision comes after months of pressure from the government, businesses, and media over the scale of ad fraud on Facebook and Instagram.
What exactly is changing
The new AI mechanism is designed to proactively detect fraud attempts and automatically block accounts impersonating celebrities, politicians, and entrepreneurs before users even get a chance to report them. According to Jakub Turowski, Meta's director of public policy for Central and Eastern Europe, an earlier version of similar technology cut reports involving the misuse of public figures' likenesses by more than 80 percent.
The second element of the changes is full identity verification for financial-sector advertisers targeting Poland. Until now, Meta mainly verified the identity of accounts flagged as suspicious. The company says the process covering all financial advertisers will be completed within the coming weeks, and that globally, by the end of 2026, 90 percent of ad revenue is expected to come from fully verified entities, up from 70 percent in 2025.
Pressure from government and business
The immediate backdrop to the announcement is a formal request by Deputy Prime Minister and Minister of Digital Affairs Krzysztof Gawkowski to the European Commission to impose a fine of around one billion zlotys on Meta for its ineffective handling of fake ads on its platforms. Gawkowski cited test results from CERT Polska, which found that Meta failed to remove as much as 86.8 percent of reported fraudulent ads.
Inaction and ineffectiveness in combating ads and fraudulent content harmful to Poles will not be tolerated. I expect the European Commission to swiftly conduct proceedings and impose a fine. There is more than enough evidence - Krzysztof Gawkowski, Deputy Prime Minister and Minister of Digital Affairs
The pressure is not limited to the government. InPost CEO Rafał Brzoska launched his own platform, ScamWatch, documenting cases of Meta's inaction on reported scams. Experts cited in Polish media estimate that Meta alone may be earning more than 20 million zlotys a day from scam ads in Poland, further fueling the dispute over the effectiveness of the company's actions so far.
Numbers that offer little reassurance
Meta touts its own statistics: an 83 percent drop in reports of fraudulent ads over two years, and the removal of 137,000 fake ads in Poland over the past year, 88 percent of which were caught automatically before anyone had a chance to report them.
These figures stand in stark contrast to the results of an independent CERT Polska test, which checked Meta's response to specific reports and found that in more than 86 percent of cases the harmful ad was not removed. The gap between the company's own statistics and the results of the external audit is exactly what the government cites as justification for its fine request.
What it means for Poland and Europe
Poland is among the first group of European Union countries to receive the upgraded version of Meta's anti-fraud system, which in itself signals how highly the company rates the reputational and regulatory risk tied to the situation in Poland. The case is unfolding in the shadow of the EU's Digital Services Act (DSA), under which the European Commission can fine Meta an amount reaching several percent of its global revenue.
For Polish businesses and users, this is above all a credibility test for Meta's new pledges. Earlier waves of fake investment ads using the likenesses of well-known figures showed how easily scammers adapt generative AI to produce convincing ad content. Whether the announced system actually curbs the scale of the problem will only become clear once the advertiser verification process, planned for the coming weeks, is complete, and after the European Commission rules on Gawkowski's request.
